The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could be used to influence agent behavior or execute unintended operations within the Zabbix environment.
Ingestion points: The script scripts/zabbix-bulk-hosts.py ingests host data from CSV files; scripts/zabbix-export.py imports configuration data from JSON files; and SKILL.md provides examples for importing templates via XML.
Boundary markers: There are no boundary markers or instructions to the agent to ignore potentially malicious instructions embedded within the data fields of the ingested files.
Capability inventory: The skill leverages the zabbix-utils library to perform powerful operations including creating, updating, and deleting hosts, triggers, and templates via the Zabbix API.
Sanitization: The provided scripts do not perform sanitization or validation on the input data before passing it to API parameters.
[CREDENTIALS_UNSAFE]: The documentation contains hardcoded default credentials for demonstration purposes.
Evidence: SKILL.md and references/api-reference.md include the line api.login(user="Admin", password="zabbix"), which refers to the well-known factory default credentials for Zabbix.

zabbix-api